S. Sierra vs S. Stephens — Prediction & Analysis

TennisEdge's AI model finds no betting value in S. Sierra vs S. Stephens (Toronto on 2026-08-03) — the market is efficient, with S. Stephens favoured at 1.73.

Model probabilities

Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.

We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?

Elo ratings

Our own ratings, computed from match results since 2018. Players with fewer than 5 rated matches are omitted rather than shown at a provisional starting value. Last updated May 29, 2026.

Head to head

S. Stephens leads 1–0 over 1 meeting. Full record.

Recent form

All Aug. 3, 2026 results & picks →

Full analysis

S. Sierra (A) vs S. Stephens (B) — WTA Toronto 1/64-finals, hard.

MODELS vs MARKET
  ELO A 48.0% | ML A 53.9% | blend A 50.2%
  Betfair ML A 2.0 / B 1.9 -> devig A 48.7% / B 51.3% (book 2.1/1.73 is stale, contradicted by exchange -> discarded).
  Models cluster on the exchange. Only ML sits above it, by ~5pp — a lone-model split; ELO and blend agree with the market.

RANKING/ELO NOTE
  Sierra ranked 72, Stephens ranked 1098 yet priced near-even — Stephens' rank massively understates her (former US Open champ, comeback). ELO (48% for the #72) treats her as a genuine peer, so the ranking gap is not a usable edge; the exchange has already priced the comeback.

RECENT FORM (last 5)
  Sierra (A): 07-27 Memphis QUAL Lost vs Prozorova (1500) 3s — UNDER-performing vs weak; 07-01 Wimb Lost vs Gauff (1701) tight TB — good shape vs elite; 06-29 Wimb Won vs Bondar (1352) GRIND; 06-22 BadHomburg Lost vs Zheng (1668) 3s; 06-21 BadHomburg FINAL Won vs Blinkova (1649) GRIND (grass title). Body of form is grass/clay; only hard-court look is the 07-27 quali loss to a 1500. Surface transition onto hard with thin/ugly hard sample = a negative, not support.
  Stephens (B): 08-01 Toronto "Final" Won vs Dart (1697) 3s [data suspect for a 1098-ranked player fresh off W100 events]; 07-30 Memphis QUAL Lost vs Vidmanova (1757); 07-28 Memphis QUAL Won vs Starodubtseva (1500) 3s; 07-23 W100 Evansville Lost; 07-22 W100 Evansville Won. Grinding ITF/qualifying — level-dropper, stats built vs weak fields.

SHAPE/PATH
  Sierra: FADING onto hard (quali loss to a 1500 as the lone hard data point), rust-lite (1 match/14d). Stephens: match-sharp but on a very heavy load; hard-court is her historical best surface, partially offsetting the fatigue.

FATIGUE (Step 2)
  B: matches_7d 3, matches_14d 5, court_time 432min, back_to_back 1, last match 08-01 (played yesterday, a 3-setter). Age 33 (>=30 -> hardened). This is an extreme spot and would warrant the -4.5pp 3-set-hangover haircut on B. BUT the market over-prices visible fatigue, and — decisively — this is a fatigue-origin tilt, not a model-vs-market tilt.
  A: matches_14d 1 (102min), rested. Age 22.

PRICING
  Effective (x0.95): A 1.95 (implied 51.3%), B 1.855 (implied 53.9%).
  A edge: ML 53.9% -> +2.6% (below 5% bar); blend 50.2% -> -1.1%; ELO 48% -> -3.3%.
  B edge: blend 49.8% -> -4.1%.
  Min odds A at 50% = 2.10; Betfair offers only 2.0 -> below min. No side clears the bar at executable prices.

GATES
  G1 edge-origin: FAIL — no side reaches the 5% numeric bar pre-narrative (best is ML-only +2.6% on A); any pro-A edge is manufactured entirely by Stephens' fatigue haircut, which G1 forbids as edge origin.
  G2 consensus:   PASS(n/a) — stake-tier only; would be lone-model at best.
  G3 fatal-risk:  n/a — ML market, no margin bet.

VERDICT: PASS. Models agree with the exchange (~48-50% A), the book's 5% "value" on A is a stale quote the Betfair price contradicts, executable A price (2.0) is below the 2.10 minimum, and the only pro-A story is fatigue — barred as an edge source. Textbook efficient market. No bet.
AH: not quoted (alt_markets handicap null).
alt_markets: none at analysis time (handicap/totals null).